[Big Trend] Start-ups developing their own LLMs in specialised fields such as law and medicine
[More business information about the start-ups featured in this article is available on Unicorn Factory’s big data platform, ‘Data Lab’.]
/Image generated by Microsoft ‘Designer’
As the government decides to develop a national flagship large-scale AI model and promotes the ‘independent AI foundation model’ project to secure technological sovereignty in AI, some start-ups that did not participate in the project are nevertheless standing out in the LLM (large language model) sector through their own technological capabilities.
The start-ups affiliated with the five teams included on the final list for the independent AI foundation model project announced by the Ministry of Science and ICT on the 4th comprise a total of 18 companies, including △Upstage △Twelve Labs △Rebellion △Liner △FuriosaAI and △Wrtn Technologies. Although they suffered the bitter disappointment of being eliminated in this evaluation, start-ups such as △Mathpresso △Tomorrow Robotics △Pebblous and △Junction Med also challenged the project through consortia, expressing confidence in their AI capabilities.
There are also ‘hidden powerhouses’ that did not apply for this project but are building differentiated technological capabilities by developing their own LLMs. Representative examples include △BHSN (law) △EveryAI Korea (medicine) △Miso (home services) △CLEVI (generative AI services) and △42Maru (lightweight LLMs).
They’ve solved even the difficult ‘law and medicine’ domains
/Graphic by Lim Jong-cheol
Legal tech start-up BHSN recently launched the law-specialised LLM ‘Alibi Astro’. The law-specialised LLM ‘Alibi Astro’, recently launched by legal tech start-up BHSN, is particularly noteworthy. Alibi Astro enhanced its expertise through CPT (continued pre-training) based on extensive legislation, case law and policy data. It was trained in the language structures and contexts used in actual legal practice through RLHF (reinforcement learning from human feedback), incorporating the opinions of experts such as lawyers.
It can understand the context of questions, infer logical relationships between documents and enable AI-based contract work, while reviewing an average 100-page English construction contract (EPC) in just one minute. It can build expert-level reasoning, including interpreting clauses and suggesting directions for revisions. Major Korean conglomerates, including CJ CheilJedang and Aekyung Chemical, currently use Alibi Astro in their day-to-day work.
EveryAI Korea's LLM specialised for the medical field, 'e1-M', is also attracting attention. Unlike existing global models that rely on English-based translation, e1 is designed to intuitively understand and reason about the context and nuances of Korean.
It scored 90.78 on 'KorMedMCQA', a medical QA benchmark developed by a KAIST research team, outperforming GPT-4o (85.61) and Claude 3.5 (86.51). It was also highly rated for its reasoning ability to comprehensively analyse patient symptoms and test results and even suggest treatment directions.
In the future, EveryAI Korea plans to develop customised LLMs for industries beyond healthcare, including legal (e1-L) and finance (e1-F), and introduce e1-M-based medical AI solutions for use by hospitals, pharmaceutical companies and insurers.
"Why use ChatGPT?"… Low-cost, high-efficiency LLMs
There are also startups challenging ChatGPT with low-cost, high-efficiency LLMs. 'Ivy', a generative AI service launched by CLEVI, is built around creating, distilling and combining small models to minimise AI development costs. It has introduced technology that efficiently manages the development process by applying automated, container-based deployment technology for training nodes (computing resources), and is characterised by significantly reducing the cost of building and operating services through efficient training pipeline design and optimisation of communication networks between nodes.
Miso, a startup providing around 200 home services, including cleaning, moving and appliance rental, is also transforming into a technology company by launching the industry's first LLM-based solution that recommends the home services customers need and supports connections with professionals. Previously, customers had to find and apply for home service categories one by one. Now, when customers consult the AI about the problems they are experiencing, it analyses them and recommends and connects them with the optimal service.
FortyTwoMaru has achieved technological differentiation with a highly reliable LLM. Its in-house developed ‘LLM42’ is a lightweight LLM optimised for Korean-language environments, characterised by its low-cost structure, high security and broad applicability in industrial settings.
Experts unanimously agree that the government should devise a strategy to make these companies, including not only those participating in the independent AI foundation model project but also other start-ups, the central pillar of the national AI strategy. Choi Seong-jin, CEO of the Startup Growth Research Institute, emphasised, “Institutional arrangements must be established so that start-ups can stand as central players in AI policy rather than mere bystanders,” adding, “The success or failure of an AI superpower depends on start-ups.”
/Graphic by Yoon Seon-jeong
Reporter Choi Tae-beom [email protected]
Source: MoneyToday (Reporter Choi Tae-beom) | https://www.mt.co.kr/future/2025/08/07/2025080518330143511
[MoneyToday start-up media platform ‘Unicorn Factory’]
<Copyright © MoneyToday, ‘Real-time News You Can See the Money In’. Unauthorised reproduction and redistribution, and use for AI training prohibited>